Growing up one of my favourite books was "The Five Biggest Ideas in Science." They are: physics' model of the atom, chemistry's periodic law, the big bang, plate tectonics, and evolution. It struck me recently that only the first two of these were the result of experimentation. That is to say they are big ideas in science, though they may not be ideas of science.
Darwin had been studying plate tectonics when he took his voyage on the HMS Beagle. That helped instill the idea that things change slowly over time. As he toured the islands of South America he made meticulous observations that came to help him develop his ideas of evolution.
I would be hard pressed to call what Darwin did 'science'. Is it science because his work was meticulous? I have known meticulous biblical scholars and historians. They are not scientists. Is it science because it applies to life/biology? The bible has something to say about where life came from but is certainly not science. Is it science because he was right? I know it is creeping into our culture that science has a monopoly on truth, but we all know that isn't so. Nor do I believe that any combination of the above qualifications are sufficient for science.
There are many ways to investigate the world. Science is one of them. Dogmatism is another. But there seems to be room for yet another. I like to call it relational analysis. That method of investigation that is rooted in observation and theory, is open to revision, but lacks the experimentation, the transparency and repeatability of science. This would include that which Darwin did on the Galapagos Islands.
Perhaps he was doing philosophy. His thoughts were a delicate balance between the empiricism of science and the rationalism of philosophy. I would like to think that science is a form of relational analysis, is a form of philosophy.
I have already contrasted it with science in that it doesn't use experimentation. If you consider the experiment as the highest quality observation, you can say relational analysis makes use of quantity.
As science is one form of relational analysis, the questions of science are a subset of the questions of relational analysis. Relational analysis can also be concerned with other relationships that do not lend themselves to scientific inquiry. "How do the books I read affect how I interact with my family?" This question is important, but far too specific to qualify for laboratory time. However, casual observations over time can reveal deep insight into this question and thus how to tweak family relationships by wisely choosing literature and an appropriate amount of reading time. Questions like "What sort of citizen make for a stable society?" are far too general to get a handle on with science. Relational analysis can help give science a starting point.
It is possible for people to not do relational analysis as it is possible for people to not do science or philosophy. Some people do not trust themselves to interpret the world around them, and leave it to others to figure out what is going on. Others believe they have figured out what is going on and are no longer revising their paradigms. Still others believe the world cannot be understood and don't try.
As one can do science, philosophy and theology poorly, one can be a poor relational analyst. A trailing indicator of a good analyst would be whether they have health, happiness and influence. The leading indicators would be the types of observations they make and how they are synthesized into theories. Do they weight history to heavily, or too lightly? Do they under or over generalize? Do they focus on non-essential elements? These are all leading indicators of a quality analyst.
While it is possible to do relational analysis rigorously it is unnecessary. Narrative is a really good tool to develop analyst skills. Internalizing narratives, fantastic, absurd or true to life, help one to grow a vocabulary of natural mechanisms that can be extrapolated to the real world and help make sense of it. And in the end that is what it is really all about.